最近用flink1.10写一个流任务，大致的逻辑，是将数据清洗以后写入hbase，中间不需要保存任何状态 这里是启动脚本： export HADOOP_CONF_DIR=/etc/hadoop/conf export HADOOP_CLASSPATH=/opt/cloudera/parcels/CDH/jars/*
/opt/flink/flink-1.10.0/bin/flink run -p 1 -m yarn-cluster -yqu root.crawler \ -ynm testRealTime22 \ -yD taskmanager.debug.memory.log=true \ -yD taskmanager.debug.memory.log-interval=10000 \ -yD yarn.nodemanager.vmem-check-enabled=false \ -yD taskmanager.memory.task.heap.size=2049m \ -yD taskmanager.memory.managed.size=512m \ -ys 1 -yjm 2048m -ytm 4096m -d \ -c com.fengjr.risk.BorrowerRelatedRiskApp /home/dev/test/real_time/risk-principal-borrower-related-1.0-beta.jar 1000000 但是，程序总是在运行的一段时间后被yarn kill，异常如下： Closing TaskExecutor connection container_e16_1571280155773_0449_01_000002 because: [2020-05-08 12:23:57.963]Container [pid=40959,containerID=container_e16_1571280155773_0449_01_000002] is running 102236160B beyond the 'PHYSICAL' memory limit. Current usage: 4.1 GB of 4 GB physical memory used; 7.6 GB of 8.4 GB virtual memory used. Killing container. Dump of the process-tree for container_e16_1571280155773_0449_01_000002 : |- PID PPID PGRPID SESSID CMD_NAME USER_MODE_TIME(MILLIS) SYSTEM_TIME(MILLIS) VMEM_USAGE(BYTES) RSSMEM_USAGE(PAGES) FULL_CMD_LINE |- 40959 40956 40959 40959 (bash) 0 1 11964416 351 /bin/bash -c /usr/java/jdk1.8.0_201-amd64/bin/java -Xmx2282749952 -Xms2282749952 -XX:MaxDirectMemorySize=462422021 -XX:MaxMetaspaceSize=100663296 -Dfile.encoding=UTF-8 -Dlog.file=/yarn/container-logs/application_1571280155773_0449/container_e16_1571280155773_0449_01_000002/taskmanager.log -Dlog4j.configuration=file:./log4j.properties org.apache.flink.yarn.YarnTaskExecutorRunner -D taskmanager.memory.framework.off-heap.size=134217728b -D taskmanager.memory.network.max=328204293b -D taskmanager.memory.network.min=328204293b -D taskmanager.memory.framework.heap.size=134217728b -D taskmanager.memory.managed.size=536870912b -D taskmanager.cpu.cores=1.0 -D taskmanager.memory.task.heap.size=2148532224b -D taskmanager.memory.task.off-heap.size=0b --configDir . -Djobmanager.rpc.address='fjr-yz-0-135' -Dweb.port='0' -Dweb.tmpdir='/tmp/flink-web-b7d8e065-7809-4918-baa1-6be5ce4f8a03' -Djobmanager.rpc.port='27863' -Drest.address='fjr-yz-0-135' -Dsecurity.kerberos.login.keytab='/yarn/nm/usercache/hbase/appcache/application_1571280155773_0449/container_e16_1571280155773_0449_01_000001/krb5.keytab' 1> /yarn/container-logs/application_1571280155773_0449/container_e16_1571280155773_0449_01_000002/taskmanager.out 2> /yarn/container-logs/application_1571280155773_0449/container_e16_1571280155773_0449_01_000002/taskmanager.err |- 326 40959 40959 40959 (java) 27897 4892 8097333248 1073185 /usr/java/jdk1.8.0_201-amd64/bin/java -Xmx2282749952 -Xms2282749952 -XX:MaxDirectMemorySize=462422021 -XX:MaxMetaspaceSize=100663296 -Dfile.encoding=UTF-8 -Dlog.file=/yarn/container-logs/application_1571280155773_0449/container_e16_1571280155773_0449_01_000002/taskmanager.log -Dlog4j.configuration=file:./log4j.properties org.apache.flink.yarn.YarnTaskExecutorRunner -D taskmanager.memory.framework.off-heap.size=134217728b -D taskmanager.memory.network.max=328204293b -D taskmanager.memory.network.min=328204293b -D taskmanager.memory.framework.heap.size=134217728b -D taskmanager.memory.managed.size=536870912b -D taskmanager.cpu.cores=1.0 -D taskmanager.memory.task.heap.size=2148532224b -D taskmanager.memory.task.off-heap.size=0b --configDir . -Djobmanager.rpc.address=fjr-yz-0-135 -Dweb.port=0 -Dweb.tmpdir=/tmp/flink-web-b7d8e065-7809-4918-baa1-6be5ce4f8a03 -Djobmanager.rpc.port=27863 -Drest.address=fjr-yz-0-135 -Dsecurity.kerberos.login.keytab=/yarn/nm/usercache/hbase/appcache/application_1571280155773_0449/container_e16_1571280155773_0449_01_000001/krb5.keytab [2020-05-08 12:23:58.020]Container killed on request. Exit code is 143 [2020-05-08 12:23:58.035]Container exited with a non-zero exit code 143. [email protected]
